JP4412339B2 - Image forming apparatus and image data correction method - Google Patents

Description

  The present invention relates to an image forming apparatus that corrects image data obtained by scanning a document with a scanner, and a correction method for the image data.

  Currently, many image forming apparatuses in which various functions such as copying, scanning, facsimile, network printing, and document server are integrated are used. In general, such an image forming apparatus is called a multifunction peripheral. Or it may be called MFP (Multi Function Peripherals).

  In recent years, documents have been stored electronically instead of paper and stored or transmitted, and there has been an increasing demand for multifunction devices that can realize document digitization by a scanner and data transmission by e-mail or the like. Yes. In recent years, multifunction peripherals have shifted from being compatible with black and white to being compatible with color. A color-compatible multifunction peripheral can read a document as it is in color and generate and transmit color image data. However, if a document of A4 size is read in full color and with a resolution of 300 dpi and the read image data is used as it is, for example, a file of a format such as PDF is generated, the file size becomes about 25 MB, Not suitable for sending. Therefore, when the read image data is transmitted, it is generally transmitted after being compressed.

  When compressing image data, if the entire image data is uniformly processed, the size can be reduced. However, there is a disadvantage that the characters included in the image data cannot be read due to the deterioration of the image quality due to the compression. May occur. If the compression rate is lowered to avoid this, the file size cannot be made sufficiently small.

  In view of this, a method of changing the compression rate for each predetermined region was considered. In this method, for example, compression is performed as follows.

  An area including the object is defined for each object such as a character or a photograph with respect to the read image data. The character area is binarized with a high resolution and is compressed by a reversible compression method such as MMR. Areas other than characters, such as photographs, are compressed at a high compression rate by an irreversible compression method such as JPEG with a reduced resolution. Using the image data obtained by the compression, a file in a format such as a PDF format is generated.

  By the way, in general, a copied original image has lower saturation than the original (original) original image, and discoloration, noise, and shakiness of lines and character edges are increased. Decreases. By duplicating, the image quality further deteriorates.

  For this reason, when a PDF file is generated by the above-described method using a copied manuscript, the image quality is noticeably deteriorated as compared with that generated using the original manuscript. Therefore, it is desired that the image quality when using a duplicate original is close to the image quality when using an original original. As a technique related to this, an image forming apparatus as in Patent Document 1 has been proposed.

According to the image forming apparatus of Patent Document 1, it is determined whether or not the original is a copy (duplication) depending on whether or not the reference pattern is printed on the original. A saturation correction condition is determined according to the determination result, and the saturation is corrected based on the determined saturation correction condition. As a result, in the case of duplication, appropriate correction for suppressing a decrease in saturation can be performed.
Japanese Patent Laid-Open No. 9-261497

  However, according to the image forming apparatus of Patent Document 1, since it is determined based on the reference pattern printed on the document whether it is an original or a copy, if a copy is created by a model that cannot print the reference pattern. , It can not be determined as a duplicate. Then, appropriate correction such as saturation correction cannot be performed, and deterioration in image quality cannot be suppressed.

  SUMMARY OF THE INVENTION In view of such problems, an object of the present invention is to make the image quality of image data obtained by reading a duplicate original more as close as possible to the image quality when an original original is read.

An image forming apparatus according to an aspect of the present invention includes an image data acquisition unit that acquires image data based on data obtained by scanning a document with a scanner, and brightness image data that generates brightness image data corresponding to the image data. An area for obtaining a black pixel inclusion area including a black pixel or a black pixel group that is a continuous black pixel, with a generation unit and a pixel on which dots are formed when the lightness image data is binarized specifying means, a deterioration in image quality value indicating the degree of deterioration of the image quality of the image data, the value associated with the small subregion number than a predetermined size from among the regions the black pixels inclusion area that certain means determined a decrease in image quality value specifying means, original value is the decrease in image quality value specifying means has determined that is the source of the image data is larger than a predetermined threshold value is a duplicate of obtaining A determining means for discriminating for, and a correcting means for performing on the image data corrected based on the result of determination by said discriminating means.

  The value related to the number includes the number or the ratio of the number to the size of the entire area of the image data.

  Preferably, the image processing apparatus includes area determination means for determining whether or not the black pixel inclusion area obtained by the area specifying means is a character area representing a character, and the image quality reduction value specifying means includes the image quality reduction value as: A density variation value indicating a degree of variation in density value in the character area is obtained based on the number of the small areas and the number of density values of pixels included in the character area, and the determination unit includes the small area. Even if the value related to the number of images is smaller than a predetermined threshold, if the density variation value is larger than the predetermined threshold, it is determined that the document is a copy.

  Preferably, when the determination unit determines that the determination unit is a duplicate, the correction unit performs a process of correcting the saturation of the image data, and the image data after the process of correcting the saturation is performed. Then, a process for removing noise is performed a predetermined number of times.

  Preferably, when the determination unit determines that the copy is a copy, the edge enhancement is performed for edge enhancement on a region other than the region representing the character of the image data after the correction by the correction unit is performed. It has a processing means.

  Preferably, when it is determined that the region representing the character of the image data after the correction by the correction unit is reversibly compressed, and the determination unit is a copy, the edge enhancement processing unit When the region other than the region representing the character of the image data after the processing for edge enhancement is irreversibly compressed, and the determination unit determines that the copy is not a copy, the correction by the correction unit is performed. Irreversible compression means for irreversibly compressing an area other than the area representing the character of the subsequent image data.

  Preferably, prior to the irreversible compression by the irreversible compression means, there is a resolution reduction means for reducing the resolution of the object of the irreversible compression.

  Note that “image data after processing for correcting saturation”, “image data after correction by correction means”, and “image data after processing for edge enhancement” are included. , Image data directly generated by the processing, and image data generated based on the generated image data (for example, binary image generated based on the generated image data) Data, image data representing an edge, or image data represented by brightness).

  According to the present invention, the image quality of image data obtained by reading a duplicate document can be made closer to the image quality when an original document is read more reliably than before.

  FIG. 1 is a diagram showing an example of a system configuration having an image forming apparatus 1 according to the present invention, and FIG. 2 is a diagram showing an example of a hardware configuration of the image forming apparatus 1.

  An image forming apparatus 1 shown in FIG. 1 is an image processing apparatus that integrates various functions such as copying, scanning, faxing, network printing, document server, and file transfer. Sometimes called a multi-function peripheral or MFP (Multi Function Peripherals).

  The image forming apparatus 1 is installed in offices or offices such as businesses, public facilities such as schools or libraries, stores such as convenience stores, and other various places and can be shared by a plurality of users. Further, it can be connected to the terminal device 2 such as a personal computer or a workstation and the server 3 via the communication line 4. As the communication line 4, the Internet, a LAN, a public line, a dedicated line, or the like is used.

  As shown in FIG. 2, the image forming apparatus 1 includes a scanner 10a, a CPU 10b, a hard disk 10c, a RAM 10d, a ROM 10e, a printing device 10f, a modem 10g, a network interface 10h, a control circuit 10k, an operation panel 10m, and the like. .

  The scanner 10a is an apparatus that optically reads an image such as a photograph, a character, a picture, or a chart drawn on a document sheet (hereinafter simply referred to as “document”) and converts it into electronic data. The read image data of the document is stored in the RAM 10d and, as will be described later, is subjected to various processes and converted into a file. In this embodiment, a color scanner is used.

  The printing device 10f is a device that prints image data read by the scanner 10a or image data transmitted from the terminal device 2 or the like on a sheet using toners of four colors of yellow, magenta, cyan, and black. is there.

  The operation panel 10m includes an operation unit and a display unit. A numeric keypad or the like is used as the operation unit, and a liquid crystal display or the like is used as the display unit. By operating the operation unit, the user gives an instruction to start or stop the processing to the image forming apparatus 1, and performs various other items such as a data destination, a scan condition, or an image file format. Can be specified. The display unit includes a screen for giving a message or an instruction to the user, a screen for inputting the type of processing and processing conditions desired by the user, and a screen showing the result of the processing executed by the image forming apparatus 1 Etc. are displayed. When a touch panel is used as the operation panel 10m, the touch panel serves as both an operation unit and a display unit. As described above, the operation panel 10m serves as a user interface for a user who operates the image forming apparatus 1.

  The CPU 10b converts the document image data read by the scanner 10a into a file of a format such as TIFF (Tagged Image File Format), JPEG (Joint Photographic Experts Group), bitmap, and PDF (Portable Document Format). I do. Further, overall control of the image forming apparatus 1 is performed, such as detection of input from the user or control of display on the operation panel 10m.

  The modem 10g has a built-in NCU (Network Control Unit), and is connected to another fax terminal via an analog public line to perform data control based on a facsimile protocol, fax data modulation / demodulation, and the like. The network interface 10h is a NIC (Network Interface Card) and connects to the terminal device 2 through a LAN or the Internet to create a protocol.

  The control circuit 10k is a circuit for controlling devices such as the hard disk 10c, the operation panel 10m, the scanner 10a, the printing device 10f, the modem 10g, and the network interface 10h.

  The hard disk 10c stores programs, data, and the like for realizing the functions of the respective units shown in FIG. These are read into the RAM 10d as necessary, and the program is executed by the CPU 10b. Some or all of these programs or data may be stored in the ROM 10e. Alternatively, some or all of the functions in FIG. 3 may be realized by hardware such as the control circuit 10k.

  Hereinafter, processing of the image forming apparatus 1 when generating a PDF file from image data read by the scanner 10a will be described.

  3 is a diagram illustrating an example of a functional configuration of the image forming apparatus 1, FIG. 4 is a flowchart for explaining a flow of processing of the preprocessing unit 14, FIG. 5 is a diagram illustrating an example of the original image GG, and FIG. Is a flowchart for explaining the flow of processing of the document discrimination unit 15, FIG. 7 is a diagram showing an example of a histogram representing the relationship between the density value and the number of pixels in the character block AKc, and FIG. 8 is a diagram illustrating the first correction processing unit 16. FIG. 9 is a diagram showing an example of a graph showing the relationship between the input value and the direct output value in gamma conversion, and FIG. 10 is a flowchart for explaining the processing flow of the photo block processing unit 17. 11 is a diagram showing an example of a histogram generated using a photograph image, FIG. 12 is a flowchart for explaining the flow of processing of the character block processing unit 18, and FIG. 13 is an example of an edge image To FIG, 14 is a diagram showing an example of a histogram used for determining a binarization threshold, FIG. 15 is a diagram for explaining the process of integration of the character string block AKR.

  As shown in FIG. 3, the image forming apparatus 1 includes a file storage unit 11, a correction condition storage unit 12, an image data acquisition unit 13, a preprocessing unit 14, a document discrimination unit 15, a first correction processing unit 16, and a photo block process. 17, a character block processing unit 18, a character block integration unit 19, a lossless compression unit 20, a second correction processing unit 21, a resolution conversion unit 22, an irreversible compression unit 23, and a file generation unit 24.

  The file storage unit 11 stores the PDF file generated by the file generation unit 24. The correction condition storage unit 12 uses conditions (correction conditions) used in correction processing by the first correction processing unit 16 and the second correction processing unit 21 for each type (“duplicate” or “original”) described later. save. This will be described in detail later.

  The image data acquisition unit 13 acquires “original image data GG” that is image data of a document read by the scanner 10a. The original image data GG is in a format such as TIFF, JPEG, or bitmap as full color image data based on R (red), G (green), and B (blue) (hereinafter referred to as “RGB”). Is output from the scanner 10a.

  Hereinafter, “data” of “image data” and “original image data” may be omitted and simply referred to as “image” and “original image”, respectively. The same applies to other image data described later.

  The preprocessing unit 14 performs preprocessing for processing after the document determination unit 15. Here, the preprocessing will be described with reference to the flowchart of FIG.

  In FIG. 4, first, HSL conversion is performed using the original image GG acquired by the image data acquisition unit 13 (# 101). As a result, an image represented by H (hue), S (saturation), and L (lightness) is generated from the RGB original image GG. Hereinafter, an image corresponding to the original image GG represented by lightness is referred to as “lightness image GL”.

  For example, when the original image GG is an image as shown in FIG. 5, a brightness image GL corresponding to the original image GG is obtained by HSL conversion as shown in FIG.

  When the expression “lightness image GL corresponding to the original image GG” is expressed, the content corresponding to the original image GG is displayed in an area having the same or corresponding width as the original image GG, as in this example. It means that it is a brightness image to represent. The same applies to other images.

  Next, binarization is performed using the generated brightness image GL, and a “binary image GV” that is a binary image corresponding to the brightness image GL is generated (# 102).

  Further, labeling processing is performed as follows using the generated binary image GV (# 103).

  First, a black pixel group, which is a continuous black pixel, is detected using a pixel in which dots are formed in the binary image GV as a black pixel. A rectangular region (rectangular region) including (enclosing) one black pixel group is defined (see AK1 and AK2 in FIG. 5). In this case, neighboring black pixel groups within a predetermined distance range may be regarded as one black pixel group, and a rectangular area (block) may be defined for the black pixel group regarded as one. In this embodiment, this method is applied. Thus, one rectangular area is defined for characters and the like that are relatively close to each other (see AK4 and AK9 in FIG. 5). In FIG. 5, rectangular areas are also defined at locations such as “second” and “third”, but are not shown.

  Further, information for identifying the defined rectangular area (hereinafter sometimes referred to as “block number”) is assigned to each rectangular area. In the following, when a rectangular area is defined, a block number is appropriately assigned. Further, for each rectangular area, block data AK indicating the position of the upper left pixel, the position of the lower right pixel, and the block number of the rectangular area is generated. In the present embodiment, the coordinates when the origin is the upper left corner of the entire image (in this case, the binary image GV) are used as the “pixel position”. The same applies hereinafter.

  Hereinafter, the block data AK may be simply referred to as “block AK”. In addition, when “block AK” is described, it may mean a rectangular area defined by the coordinates shown there. The same applies to other block data AK described later.

  Also, the block AK may be described separately from “block AK1”, “block AK2”,. The same applies to the other symbols.

  As shown in FIG. 5, by this labeling process, a block AK is defined for each object such as a character, a photograph, a figure, a graph, and a table included in the binary image GV.

  The brightness image GL, the binary image GV, and the block data AK generated here are appropriately used as necessary in the subsequent processing.

  The document discriminating unit 15 performs a process for discriminating whether the original document from which the original image GG acquired by the image data acquiring unit 13 is copied or original.

  In this process, when it is determined that there is a deterioration in image quality in the original document due to the amount of noise (noise amount) included in the original image GG and the degree of blur (degree or state) of the original image GG, The original is determined to be a duplicate. Here, the processing of the document discrimination section 15 will be described in detail with reference to the flowchart of FIG.

  In FIG. 6, first, discrimination based on the amount of noise is performed. At that time, the block size of each block AK (size based on the number of vertical and horizontal pixels of the block) is obtained from the coordinates indicated in the block data AK. Based on the block size, a minute block AK is specified. Specifically, a block AK whose block size is smaller than a predetermined size (for example, vertical × horizontal = 100 × 100 pixels) is specified as a minute block (for example, blocks AK11 to AK14 in FIG. 5) ( # 111). Hereinafter, the minute block AK specified here may be referred to as an “isolated point block AKk”.

  Next, the number of identified isolated point blocks AKk is counted (# 112). If the number is larger than a predetermined threshold (for example, 20 or 30) (Yes in # 113), the original document is assumed to contain a lot of noise, and the document is “duplicated”. (# 114). Alternatively, the ratio of the number of isolated point blocks AKk to the size of the original image GG (the size based on the number of vertical and horizontal pixels of the document) is obtained, and when the ratio is larger than a predetermined threshold value, “duplicate” is determined. May be.

  Normally, a duplicated document contains more noise than the original document due to toner dust and the like. Therefore, by detecting the amount of noise as described above, the type of document, that is, “duplicate” or “original” can be determined.

  If the number of isolated point blocks AKk is equal to or smaller than a predetermined threshold (No in # 113), next, a determination is made based on how the original image GG is blurred.

  At that time, first, a block AK (character block AKc) that can be determined to be defined for a character is specified (# 115). At this time, for example, a range of the size of the block AK that includes one character is determined in advance, and the block AK having a size within the range is determined as the character block AKc.

  Or you may determine by the ratio (binarization pixel rate) with respect to the size of the block AK of the number of the black pixels which are the pixels by which the dot is struck in the part corresponding to the block AK of the binary image GV.

  In the case of a block AK defined for a character, some margin is included between the character and the circumscribed rectangle of the block AK. In consideration of the margin, when the binarized pixel rate is equal to or less than a predetermined threshold, the block AK is determined as the character block AKc. Note that when the size of the block AK is within a predetermined range and the binarized pixel ratio is equal to or less than a predetermined threshold, the block AK may be determined as the character block AKc.

  Next, the degree (diffusion degree) of lightness diffusion (variation) within the specified character block AKc is obtained. The degree of blurring of the original image GG is detected based on the degree of diffusion.

  In this case, first, using a portion corresponding to the character block AKc of the lightness image GL, the relationship between the density value (density value in the lightness image) and the number of pixels as shown in FIG. 7A is shown. A histogram is generated (# 116).

  In FIG. 7A, the horizontal axis indicates the density value of the pixel, and the vertical axis indicates the number of pixels. Each column of the histogram indicates the number of pixels of each density value in the character block AKc.

  A peak MT having an extreme value greater than a predetermined threshold value (number) VK is detected in the generated histogram. Then, a width WT at a predetermined height (number) VL of the mountain MT is obtained. Such a width WT may be obtained for all the character blocks AKc included in the binary image GV, or may be obtained for a predetermined number of character blocks AKc.

  An average value of the respective widths WT obtained for each character block AKc is obtained, and is set as the diffusion degree (# 117). That is, the degree of diffusion of the histogram with respect to the peak PK1 of the mountain MT is defined as the diffusion degree. Note that one character block AKc may be selected, the width WT for that character block AKc may be obtained, and this may be used as the diffusivity.

  If the degree of diffusion is greater than a predetermined threshold (Yes in # 118), the original document is determined to be “duplicate” (# 114). If the degree of diffusion is equal to or less than the predetermined threshold (No in # 118), the original document is determined as “original” (# 119).

  In a copied manuscript, an edge portion of an image included in the manuscript is often blurred or struck. Usually, since the characters are represented by a single color, it is easy to compare the edge portion between the image with much blurring and the image with little blurring depending on the variation of the density value in the character block AKc. That is, when a histogram is generated by the above-described method when a lot of blurs are included, for example, as shown in FIG. 7B, the width WT is larger than a histogram without blurs (for example, FIG. 7A). Is relatively long. By using such features, the type of document can be determined.

  The first correction processing unit 16 performs correction on the original image GG based on the determination result by the document determination unit 15. This process will be described with reference to the flowchart of FIG.

  In FIG. 8, when the result of determination by the document determination unit 15 is “duplicate” (Yes in # 131), the correction condition for the copy is acquired from the correction condition storage unit 12 (# 132).

  Based on the acquired correction conditions, brightness correction is performed on the original image GG, and the original image GG is corrected in a brighter direction with respect to brightness (# 133). Hereinafter, the original image GG subjected to the brightness correction may be referred to as “original image GGb”. For this correction method, for example, a known gamma conversion is used.

  In gamma conversion, the density value of each pixel of the original image GG is corrected using the following equation (1). Specifically, the density value of each pixel of the original image GG is substituted as an input value (x) into Expression (1) to obtain an output value (y) that is a density value after correction.

  As shown in FIG. 9, the output characteristic with respect to the input value can be changed by changing the gamma value (γ) to various values. In the case of using gamma conversion, a gamma value that can correct the original image GG obtained from the copy original to an appropriate brightness is stored in advance in the correction condition storage unit 12 as a correction condition for copying.

  Next, noise removal processing (hereinafter simply referred to as “noise removal”) based on the correction condition acquired in step # 132 is performed on the generated original image GGb (# 134). For example, a known median filter is used as a noise removal method. By using the median filter, noise such as isolated points can be removed without blurring the edge portion of the image.

  In many cases, the effect of noise removal can be enhanced by repeating the noise removal on the image after the noise removal.

  In the present embodiment, when the determination result is “replication”, this noise removal processing is repeated, for example, 10 times. For noise removal, this number of times is stored in the correction condition storage unit 12 as a correction condition.

  The reason for removing noise from the original image GGb is to make a beautiful image appear in a PDF file generated later.

  In step # 134, the same noise removal (10 times) is performed on the brightness image GL. The brightness image from which noise has been removed is used for character detection in a character block processing unit 18 described later. By removing the noise, it is possible to reduce the erroneous detection of noise as a character and to increase the detection accuracy when detecting the character. Thereby, not only can the image be cleaned, but also the effect of compression by generating a PDF file can be enhanced.

  If the result of determination by the document determination unit 15 is “original” (No in # 131), the noise removal correction condition for the original is acquired from the correction condition storage unit 12 (# 135). Noise removal is performed on the original image GG for the number of times indicated by the acquired correction condition (# 134). Similarly, noise removal is performed on the brightness image GL.

  In the case of “original”, a smaller number of times (for example, “once” etc.) than in the case of “duplicate” is stored as a correction condition. It may be “0 times”.

  Normally, an original document has a smaller amount of noise than a duplicate document, and thus, the number of times of noise removal is made smaller than that in the case of duplication and priority is given to improving processing efficiency.

  Since the reversible compression unit 20 and the irreversible compression unit 23, which will be described later, perform compression applying run length coding such as MMR (Modified Modified READ) and JPEG, the effect of the compression can be enhanced by reducing noise. .

  Hereinafter, the original image GG or the original image GGb subjected to the noise removal process in Step # 134 is referred to as “original image GGc”, and the lightness image GL subjected to the noise removal is referred to as “lightness image GLb”. is there. The original image GG, the original image GGb, and the original image GGc may be collectively referred to as “original image GG”. Similarly, other symbols may be collectively described by omitting serial numbers attached to the symbols. The same applies to the drawings.

  The photo block processing unit 17 performs processing for specifying a block AK (photo block AKp) that can be determined to be defined for a photo portion. This process will be described with reference to the flowchart of FIG.

  In FIG. 10, first, by identifying a block AK (frame line block AKw) and a character block AKc that can be determined to be defined for the frame line, other blocks AK are extracted as candidates for the photo block AKp (# 141). ).

  The character block AKc is specified by the method already described in the process of step # 115 of the document discrimination unit 15. Alternatively, the block number of the block AK determined as the character block AKc by the process of step # 115 of the document determination unit 15 may be acquired and specified.

  The frame line block AKw can also be specified using the block size or the binarized pixel rate in the same manner as the character block AKc.

  Here, when specifying the character block AKc and the frame line block AKw, the lightness image GL may be used as in the case of the document determination unit 15, or the lightness image GLb may be used.

  Next, the photograph block AKp is identified from the blocks AK extracted as candidates for the photograph block AKp (# 142).

  Usually, a photo contains more colors than text. Therefore, when a histogram representing the relationship between the density value and the number as described in FIG. 7 is created, a histogram over a wide range of density values is generated as shown in FIG. Using this feature, for example, when a width WTb of a mountain MTb having an extreme value equal to or greater than a predetermined threshold VM is obtained at the height (number) of the threshold VM, and the width WTb is equal to or greater than a predetermined length Is determined as a photo block AKp. Alternatively, when there is no mountain whose extreme value exceeds the predetermined threshold value VN, it is determined as a photo block AKp.

  In the determination for the photographic block AKp, a histogram for each of RGB is generated using the original image GGc. The original images GG and GGb may be used. Further, it may be determined as a photo block AKp when the above conditions are satisfied in each of the RGB histograms, or may be determined as a photo block AKp when any of the histograms is satisfied. Alternatively, a histogram obtained by combining the histograms of RGB may be generated, and the photo block AKp may be determined when the above conditions are satisfied in the histogram.

  The character block processing unit 18 performs a process for separating a character portion included in the original image GG from other portions. Here, a part other than the photo block AKp specified by the photo block processing unit 17 is set as a processing target. This is because when trying to detect a character included in a photograph, a character that is not actually a character may be erroneously detected as a character.

  Therefore, first, a portion corresponding to the photo block AKp is removed from the lightness image GLb (or lightness image GL). At this time, for example, a portion corresponding to the photo block AKp is painted in white or black. Or you may make it handle as the thing which does not have the pixel of the part. Hereinafter, the lightness image GLb (or lightness image GL) from which the part corresponding to the photo block AKp has been removed may be referred to as “lightness image GLc”.

  Here, the processing of the character block processing unit 18 will be described with reference to the flowchart of FIG.

  In FIG. 12, first, an edge portion is extracted from the brightness image GLb to generate an “edge image GE” that is an edge image corresponding to the brightness image GLc (# 151).

  An edge image is an image representing an edge portion as shown in FIG. 13, for example. FIG. 13A shows an edge image generated based on an image of a building photograph. FIG. 13B shows an edge image generated based on an image representing a table.

  In the example of FIG. 13, the edge portion is represented by white pixels (white pixels) on which dots are not applied, and the other portions are represented by black pixels. Note that the edge portion may be represented by black pixels and the other portions may be represented by white pixels.

  Next, in order to improve the accuracy of ruled line detection in step # 153, a process of correcting the inclination (inclination correction) is performed on the generated edge image GE and brightness image GLc (# 152). Various known methods can be used for this inclination correction. The inclination correction may be performed on the entire image or may be performed for each block AK.

  Hereinafter, the edge image GE subjected to inclination correction may be referred to as “edge image GEb”, and the lightness image GLc subjected to inclination correction may be referred to as “lightness image GLd”.

  Although a binary image may be used here, an edge image is used here because the use of the edge image has an advantage that the processing can be performed faster than the binary image.

  Next, a ruled line is detected using the brightness image GLd (# 153). This processing is performed to remove the ruled line portion from the character detection processing target in the subsequent step # 154. By removing the ruled line portion, the accuracy of character detection can be increased.

  First, when pixels having a density value equal to or greater than a predetermined threshold value (density value in the brightness image) in the lightness image GLd are continuous (continuous) for a predetermined length or longer, the continuous pixel group is detected as a ruled line. The ruled line portion is removed from the brightness image GLd. Also, the coordinates of the pixel group are acquired. Based on the coordinates, the ruled line portion is also removed from the edge image GEb. Hereinafter, the lightness image GLd and the edge image GEb from which the ruled lines have been removed may be referred to as “lightness image GLe” and “edge image GEc”, respectively.

  Next, a process for specifying the portion where the character is represented using the edge image GEc is performed (# 154). This process is performed as follows. In the edge image GEc, it is assumed that the edge portion is represented by white pixels as shown in FIG.

  First, in the edge image GEc, black pixels that are continuous for a predetermined length in the horizontal direction (horizontal direction) are detected. In many cases, black pixels are continuous in a portion between character lines. Therefore, such a portion is determined to be a line spacing and excluded from the processing target, and the other portion is extracted as the processing target.

  A character portion is further specified from the extracted portion. In the edge image GEc, the character portion is composed of continuous white pixels (white pixel group). Therefore, first, such a white pixel group is detected. Further, neighboring white pixel groups within a predetermined distance range in the horizontal direction are grouped together as the same group, and a rectangular region including one group is defined. Row block data AKb indicating the position of the upper left pixel, the position of the lower right pixel, and the block number of the defined rectangular area is generated.

  If the line block AKb is defined for a character string, it should have a shape extending in the horizontal direction. Therefore, the ratio (vertical length / horizontal length) between the vertical length and the horizontal length of the defined row block AKb is obtained, and when the ratio is less than a predetermined threshold, the row block AKb is It is determined that it is defined for the character string.

  Alternatively, the determination may be made based on the ratio of the number of white pixels to the size of the row block AKb (in-block white pixel ratio). That is, when the in-block white pixel ratio is equal to or less than a predetermined threshold, it is determined that the line block AKb is defined for the character string. Hereinafter, the row block AKb determined to be defined for the character string may be referred to as a “character string block AKr”. Note that the character string block AKr may be defined for one character instead of a character string.

  Next, a process for specifying pixels forming a character part in the character string block AKr is performed (# 155). In this processing, the portion corresponding to the character string block AKr of the lightness image GLe (or lightness image GLd) is binarized based on a predetermined threshold value, and the pixel with dots is identified as the pixel forming the character. To do. Then, the position of the pixel is obtained.

  Here, the threshold for binarization is determined for each character string block AKr. The determination is performed as follows, for example.

  First, a histogram showing the relationship between the density value of pixels in the character string block AKr (density value in the brightness image) and the number of pixels is generated (see FIG. 14). The height HPb of a predetermined ratio with respect to the height HP of the peak PK2 of the generated histogram is obtained. Further, the density value VP of the histogram at the height HPb is obtained and used as a threshold value for binarization. In this way, a threshold value is obtained for each character string block AKr. In binarization, for example, dots are applied to pixels whose density value is larger than the threshold value. The pixel on which the dot is placed is determined as a pixel forming a character, and its position is obtained. As a result, the character portion and the background portion can be separated. In addition, a binary image for each character string block AKr obtained by binarization (hereinafter sometimes referred to as “block binary image BV”) is added to the corresponding character string block data AKr. The block binary image BV obtained here is used in the subsequent compression by the lossless compression unit 20.

  Next, the color of the character (character color) is specified by referring to the RGB density values of the pixels of the original image GG corresponding to the obtained pixels at each position (# 156). Data indicating the specified character color (RGB density value) is added to the character string block data AKr.

  As the character color, an average value of density values of each pixel may be used, or a maximum value or a minimum value may be used. In specifying the character color, the original image GGb or GGc may be used.

  The character block integration unit 19 performs processing for integrating the character string blocks AKr that have a relationship satisfying a predetermined condition into one block. In the present embodiment, those having similar character colors specified by the character block processing unit 18 are integrated. That is, those having character color differences (character color differences) within a predetermined range are integrated. The character color difference can be obtained by, for example, the following formula (2). In Expression (2), “R0”, “G0”, and “B0” are the density values of RGB of the character color of a certain character string block AKr, “R1”, “R2”, and “B3”, respectively. "Indicates each density value of RGB of the character color of the other character string block AKr.

  Alternatively, the same character color may be integrated.

  Here, the integration process will be specifically described with reference to FIG.

  In FIG. 15A, a plurality of character string blocks AKr are arranged at positions corresponding to the respective coordinates in an area AR having the same area as the original image GG. Here, when the character color difference between the character string block AKr2 and the character string block AKr3 is within a predetermined range, as shown in FIG. 15B, they are integrated to generate one character string block AKr9. That is, the character string block data AKr9 indicating the position of the upper left pixel PT1, the position of the lower right pixel PT2 and the block number of the rectangular area is generated.

  The average value of the character color of the character string block AKr2 and the character color of the character string block AKr3 is set as the character color of the character string block AKr9.

  Similarly, the other character string blocks AKr that satisfy the conditions are integrated. The character string block AKr formed by the integration is added as an integration target, and such integration is repeated until there are no combinations that satisfy the condition.

  In a PDF file, one color is handled as one layer. Therefore, the size of the generated PDF file can be reduced by collecting characters of similar colors as one color.

  The lossless compression unit 20 compresses the block binary image BV corresponding to the character string block AKr remaining after the integration process of the character block integration unit 19 by the lossless compression method. Data indicating the character color of the character string block AKr is added to the data generated by the compression to generate the compressed character image data DAc. The compressed character image data DAc is generated for each character string block AKr.

  In this embodiment, MMR is used as a lossless compression method. Since the outline of the character becomes clear by the MMR processing, edge enhancement is not performed on the character string block AKr.

  The second correction processing unit 21 performs processing for emphasizing edges of portions other than the character string block AKr in the original image GGc when the result of determination by the document determination unit 15 is “duplicate”. The parameters for edge enhancement processing are stored in the correction condition storage unit 12 in advance.

  In the present embodiment, edge enhancement is performed by a known method using a sharpening filter such as a Laplacian filter.

  The resolution conversion unit 22 performs a process for reducing the resolution of a portion other than the character string block AKr in the original image GGc.

  The irreversible compression unit 23 compresses the reduced resolution portion using a irreversible compression method. Thus, compressed non-character image data DAh is generated. In the present embodiment, JPEG is used as a lossy compression method.

  The file generation unit 24 appropriately processes the generated compressed character image data DAc and compressed non-character image data DAh by adding necessary data to generate a PDF layer. A PDF file DF is generated by the layer and stored in the file storage unit 11.

  FIG. 16 is a flowchart for explaining an example of a processing flow of the image forming apparatus 1 when generating a PDF file. Next, the processing flow of the image forming apparatus 1 when the original is read by the scanner 10a and saved as a PDF file will be described with reference to the flowchart of FIG.

  When generating a PDF file, the user first sets scanning conditions such as resolution using the operation panel 10m. Further, PDF is designated as the image storage format.

  When the user instructs execution of scanning through the operation panel 10m, the image forming apparatus 1 scans the document with the scanner 10a (# 201).

  A color image (RGB image) obtained by scanning is subjected to HSL conversion to generate a brightness image (# 202). Further, a binary image is generated by binarizing the brightness image (# 203). By performing a labeling process on the generated binary image, a rectangular region including a black pixel group included in the binary image is defined (# 204).

  Next, the noise amount is obtained using the binary image. Further, a rectangular region representing the character is specified, and the diffusion degree of the density value of the portion corresponding to the rectangular region in the brightness image is obtained (# 205).

  The amount of noise is determined by determining a minute one of the rectangular areas defined in step # 204 as noise and counting the number of the noises. Further, the ratio of the number to the entire image may be the amount of noise. In obtaining the diffusivity, first, a rectangular area that can be determined to be defined for a character is identified from the ratio of black pixels to the size of the rectangular area in the binary image. Then, a portion corresponding to the rectangular area in the brightness image is extracted. A histogram representing the density value and the number of pixels in each density value is generated for the extracted portion. The width of the peak in the histogram is obtained, and the width is defined as the diffusivity.

  If the obtained noise amount is larger than a predetermined threshold value or the diffusivity is larger than the predetermined threshold value, it is determined that the read original is a duplicate (# 206). If this condition is not satisfied, it is determined that the document is original.

  Next, correction based on the determination result is performed (# 207). When the determination result is “duplicate”, brightness correction is performed on the read image. Further, the noise removal process is performed “10 times” for each of the image subjected to the brightness correction and the brightness image generated in step # 202. When the determination result is “original”, noise removal is performed “once” for each of the read image and the brightness image.

  Next, using the corrected image, the rectangular area defined in step # 204 that can be determined to be defined for the portion of the photograph (rectangular area for the photograph) is specified (# 208). Many colors are usually used for the image of the photograph. Here, a rectangular area is specified using the feature.

  Next, the portion of the rectangular area specified as defined for the photograph is excluded from the corrected brightness image, and the character portion is specified from the remaining portion (# 209).

  In this case, first, an edge portion of the brightness image is extracted to generate an edge image. An inclination correction process is performed on the generated edge image and brightness image. Further, in the edge image, when black pixels forming an edge are continuous for a predetermined length or longer, these pixels are determined to be ruled lines and removed. Corresponding pixels are also removed from the brightness image. A white pixel group of white pixels (pixels representing an edge portion) continued from the remaining portion in the edge image is detected, and a rectangular region including the white pixel group is defined. Whether or not the white pixel group represents a character is determined from the ratio of the vertical and horizontal lengths of the rectangular area. By this determination, the character portion is specified.

  Next, for each rectangular area (rectangular area for a character) for a white pixel group determined to represent a character, the color of the character shown there is specified. In this case, first, in the brightness image from which the ruled line portion has been removed, the portion corresponding to the rectangular region is binarized. And each pixel by which the dot was hit is specified as a pixel which forms a character. The RGB density values of the pixels corresponding to the specified pixels in the RGB image after the correction processing in step # 207 are acquired. The average value of the acquired density values is used as the color of the character represented in the rectangular area.

  If the result of determination in step # 206 is “duplicate” (Yes in # 210), edge enhancement processing is performed on the RGB image after correction in # 207 for the portion other than the rectangular area for the character ( # 211). The resolution of the edge-enhanced portion is reduced (# 212). Further, the portion is compressed by the JPEG method (# 213). If the result of determination in step # 206 is “original”, the processing of steps # 212 and # 213 is performed without performing the edge enhancement processing of step # 211.

  Further, among the rectangular areas for the characters, those having similar character colors are integrated (# 214). Further, the binary image generated when specifying the character color corresponding to the rectangular area remaining after the integration is compressed by the MMR method (# 215).

  A PDF file is generated using the compressed data generated by the compression in step # 213 and step # 215 (# 216).

  Normally, when a document is a duplicate, the image quality obtained by scanning it is worse than that of an image obtained from the original document. According to the present embodiment, appropriate correction is performed according to the type (“duplicate” or “original”) of the original document that is the source of the input image. The image quality closer to the original can be realized than the image quality that had been used.

  In addition, according to the present embodiment, since such a type is automatically determined based on the characteristics of the input image, it is possible to easily obtain an image with higher image quality than before.

  In addition, since the amount of noise can be effectively reduced at the stage before compression, the compression effect can be enhanced.

  When specifying a character part, if the character part is specified as a part other than the character, the readability of the character deteriorates due to low resolution and lossy compression. If a portion other than characters is specified as a character portion, the compression rate is lowered. According to the present embodiment, the accuracy to be specified can be increased by reducing the amount of noise, so that it is possible to improve the image readability and the compression rate as well as the image quality.

  In the present embodiment, the width of the peak of the histogram is used as the degree of diffusion when determining the type of document. However, the standard deviation of density values may be obtained and used as the degree of diffusion.

  In this case, the portion corresponding to the character block AKc of the brightness image GL is extracted, the average value of the density values of the pixels included therein and the total number of pixels are obtained, and the standard deviation is obtained using the following equation (3). .

  In this embodiment, the image forming apparatus 1 automatically determines the type of the document. However, the user may be able to specify whether the document is a copy.

  In the present embodiment, the number of times (degree of correction) for performing saturation correction and further removing noise is changed according to the type of document. Alternatively, the correction method may be changed according to the type of document.

  In the present embodiment, an example in which brightness correction is performed has been described. However, saturation correction for correcting the saturation in a vivid direction may be performed. In this case, a similar method such as gamma conversion can be used.

  Further, although an example in which the brightness correction is performed only in the case of duplication has been shown, it may be performed in the case of the original. In that case, the gamma value may be changed according to the type of document.

  In addition, the configuration, functions, threshold values, contents indicated by each data, contents or order of processing, and the like of the whole or each part of the image forming apparatus 1 can be appropriately changed in accordance with the spirit of the present invention.

1 is a diagram illustrating an example of a system configuration having an image forming apparatus according to the present invention. 2 is a diagram illustrating an example of a hardware configuration of an image forming apparatus. FIG. 2 is a diagram illustrating an example of a functional configuration of an image forming apparatus. FIG. It is a flowchart for demonstrating the flow of a process of a pre-processing part. It is a figure which shows the example of an original image. 5 is a flowchart for explaining a flow of processing of a document discrimination unit. It is a figure which shows the example of the histogram showing the relationship between the density | concentration and the number of the pixels in a character block. It is a flowchart for demonstrating the flow of a process of a 1st correction process part. It is a figure which shows the example of the graph showing the relationship between the input value and output value in gamma conversion. It is a flowchart for demonstrating the flow of a process of a photograph block process part. It is a figure which shows the example of the histogram produced | generated using the image of the photograph. It is a flowchart for demonstrating the flow of a process of a character block process part. It is a figure which shows the example of an edge image. It is a figure which shows the example of the histogram used when determining the threshold value of binarization. It is a figure for demonstrating the process of integration of a character string block. 10 is a flowchart for explaining an example of a processing flow of the image forming apparatus when generating a PDF file.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Image forming apparatus 13 Image data acquisition part (image data acquisition means)
14 Pre-processing part (area specifying means)
15 Document discriminating section (image quality degradation value specifying means, discriminating means, area discriminating means)
16 1st correction process part (correction means)
20 Lossless compression unit (reversible compression means)
21 2nd correction process part (edge emphasis means)
22 Resolution converter (resolution reduction means)
23 Lossy compression section (lossy compression means)
AK block (black pixel inclusion area)
AKc character block (character area)

Claims (13)

Image data acquisition means for acquiring image data based on data obtained by scanning a document with a scanner;
Brightness image data generation means for generating brightness image data corresponding to the image data;
A region specifying means for determining a black pixel inclusion region including a black pixel group that is a black pixel as a pixel in which dots are formed when the lightness image data is binarized;
An image quality reduction value for obtaining a value related to the number of small areas smaller than a predetermined size from the black pixel inclusion area obtained by the area specifying unit as an image quality reduction value indicating a degree of image quality degradation of the image data. Specific means,
A determination unit that document the value is the decrease in image quality value specifying means has determined that is the source of the image data is larger than a predetermined threshold value is determined to be a duplicate,
Correction means for performing correction on the image data based on the determination result by the determination means;
An image forming apparatus comprising: Image data acquisition means for acquiring image data based on data obtained by scanning a document with a scanner;
A region specifying means for determining a black pixel inclusion region including a black pixel group that is a black pixel, or a black pixel group that is a continuous black pixel, with a pixel to which a dot is formed when the image data is binarized as a black pixel;
Area determining means for determining whether or not the black pixel inclusion area obtained by the area specifying means is a character area representing a character ;
The character area together with a value related to the number of small areas smaller than a predetermined size from the black pixel inclusion area obtained by the area specifying means as an image quality deterioration value indicating a degree of deterioration of the image quality of the image data a decrease in image quality value specifying means asking you to density variation value indicating a degree of variations in the density value of the character area on the basis of the number of density values of pixels included in,
When the value related to the number of small areas obtained by the image quality degradation value specifying means is greater than a predetermined threshold, it is determined that the original document from which the image data is based is a copy, and the value related to the number of small areas Discriminating means for discriminating that the document is a copy when the density variation value is larger than the predetermined threshold even when the value to be performed is smaller than the predetermined threshold ;
Correction means for performing correction on the image data based on the determination result by the determination means;
An image forming apparatus comprising: Image data acquisition means for acquiring image data based on data obtained by scanning a document with a scanner;
A region specifying means for determining a black pixel inclusion region including a black pixel group that is a black pixel, or a black pixel group that is a continuous black pixel, with a pixel to which a dot is formed when the image data is binarized as a black pixel;
Area determining means for determining whether or not the black pixel inclusion area obtained by the area specifying means is a character area representing a character ;
A density variation value indicating the degree of variation in density value in the character area is obtained based on the number of density values of pixels included in the character area as an image quality degradation value indicating the degree of degradation in image quality of the image data. Image quality degradation value specifying means,
If before Symbol density variation value is greater than a predetermined threshold, the determination means document is the source of the image data is determined to be a duplicate,
Correction means for performing correction on the image data based on the determination result by the determination means;
An image forming apparatus comprising: The correction unit performs a process of correcting the saturation of the image data when the determination unit determines that the copy is a copy.
The image forming apparatus according to any one of claims 1 to 3. The correction means, when it is determined that the determination means is a copy, performs a process for removing noise of the image data a predetermined number of times,
The image forming apparatus according to any one of claims 1 to 3. The correction means performs a process for correcting the saturation of the image data when the determination means determines that the copy is a duplicate, and performs noise processing on the image data after the process for correcting the saturation. Perform a predetermined number of times to remove
The image forming apparatus according to any one of claims 1 to 3. Reversible compression means for reversibly compressing an area representing characters of image data after correction by the correction means;
Irreversible compression means for irreversibly compressing an area other than an area representing characters of image data after correction by the correction means;
The image forming apparatus according to any one of claims 1 to 6 having a. Edge enhancement processing means for performing edge enhancement processing on an area other than an area representing characters of image data after correction by the correction means when the determination means determines that the copy is a copy; Have
The image forming apparatus according to any one of claims 1 to 6. Reversible compression means for reversibly compressing an area representing characters of image data after correction by the correction means;
If the determination unit determines that the copy is a copy, the region other than the region representing the character of the image data after the processing for edge enhancement by the edge enhancement processing unit is irreversibly compressed, and the determination If it is determined that the means is not a copy, an irreversible compression means for irreversibly compressing an area other than the area representing the characters of the image data after the correction by the correction means;
The image forming apparatus according to claim 8 . Prior to the irreversible compression by the irreversible compression means, the image processing apparatus has a resolution reduction means for reducing the resolution of the object of the irreversible compression.
The image forming apparatus according to claim 7 or 9 . A method for correcting image data in an image forming apparatus that has acquired image data based on data obtained by scanning a document with a scanner,
A brightness image data generation step for generating brightness image data corresponding to the image data;
An area specifying step for obtaining a black pixel inclusion area including a black pixel group that is a black pixel or a black pixel group that is a continuous black pixel, assuming that a pixel on which dots are formed when the lightness image data is binarized is a black pixel;
An image quality reduction value for obtaining a value related to the number of small areas smaller than a predetermined size from the black pixel inclusion area obtained by the area specifying unit as an image quality reduction value indicating a degree of image quality degradation of the image data. Specific steps,
A determining step of original values determined in the degradation of image quality value specifying step becomes the original of the image data is larger than a predetermined threshold value is determined to be a duplicate,
A correction step for performing correction on the image data based on the determination result in the determination step;
A correction method characterized by comprising: A method for correcting image data in an image forming apparatus that has acquired image data based on data obtained by scanning a document with a scanner,
A region specifying step for obtaining a black pixel inclusion region including a black pixel group that is a black pixel or a black pixel group that is a continuous black pixel, assuming that a pixel on which dots are formed when the image data is binarized is a black pixel;
An area determination step of determining whether or not the black pixel inclusion area obtained in the area specifying step is a character area representing a character;
The character region as well as a value related to the number of small regions smaller than a predetermined size from the black pixel inclusion region obtained in the region specifying step as an image quality degradation value indicating a degree of degradation in image quality of the image data An image quality lowering value specifying step for obtaining a density variation value indicating a degree of density value variation in the character area based on the number of pixels included in each density value;
When the value related to the number of small areas obtained in the image quality reduction value specifying step is larger than a predetermined threshold, it is determined that the document from which the image data is based is a copy, and the value related to the number of small areas A determination step of determining that the document is a copy if the density variation value is greater than the predetermined threshold even if the value to be performed is smaller than the predetermined threshold;
A correction step for performing correction on the image data based on the determination result in the determination step;
A correction method characterized by comprising: A method for correcting image data in an image forming apparatus that has acquired image data based on data obtained by scanning a document with a scanner,
A region specifying step for obtaining a black pixel inclusion region including a black pixel group that is a black pixel or a black pixel group that is a continuous black pixel, assuming that a pixel on which dots are formed when the image data is binarized is a black pixel;
An area determination step of determining whether or not the black pixel inclusion area obtained in the area specifying step is a character area representing a character;
A density variation value indicating a degree of variation in density value in the character area is obtained as an image quality degradation value indicating a degree of degradation in image quality of the image data based on the number of pixels in each density value in the character area. An image quality degradation value identification step;
A discriminating step for discriminating that the original document of the image data is a copy when the density variation value is larger than a predetermined threshold;
A correction step for performing correction on the image data based on the determination result in the determination step;
A correction method characterized by comprising:

Images